Children Fare Better In Traditional Families Than Gay Families

[ERayR]

Zeta, I don't believe the APA studies were done for political purposes, although, I realize, they are used that way, at times.

I think this new study was done for political purposes, because it was funded by a conservative group (was my understanding). The movement for GLBT rights has been making a lot of inroads, lately, and I think this study was an attempt to try and throw a wrench in the works.

And you base this on what evidence?

[Zeta-Flux]

ERayR,

I've asked libs now, multiple times, for evidence for her assertions. Nothing is coming.

And she doesn't realize that many of the studies done which were quoted by the APA were funded by pro-LGBT groups. (Nor that the APA itself has become political.)

[ERayR]

ERayR,

I've asked libs now, multiple times, for evidence for her assertions. Nothing is coming.

And she doesn't realize that many of the studies done which were quoted by the APA were funded by pro-LGBT groups. (Nor that the APA itself has become political.)

Tunnel Vision?

[Zeta-Flux]

Yup. It's all a smokescreen anyway. The study itself is strong evidence that it is rational to question the validity of previous studies and their strong claims, or at least how they are used in politics and case-law. It will prevent more misuse in the political arena than cause it.

[Libs]

And you base this on what evidence?

The new study, which was funded by conservative groups such as the Witherspoon Institute and the Bradley Foundation, has drawn national media attention because it was performed by Mark Regnerus, a sociology professor at the University of Texas-Austin whose qualifications outstrip previous "experts" touted by same-sex marriage opponents. Opponents of same-sex marriage have had great success exploiting conservative religious and cultural attitudes towards homosexuality in referendums and ballot measures. But in they've struggled in court, where they've lacked reputable social science justifications for their views on same-sex marriage. When compelled to prove that withholding marriage rights from two consenting adults of the same gender is a legitimate government interest, they've been armed with little more than their own assumptions.

http://www.motherjon...g-mark-regnerus

[california boy]

What is strange is that you seem to be claiming that a random sample was a negative aspect of this study. To the contrary, it is what makes this study stand out. True, it means there are limitations on what can be claimed. But the other studies suffer from the lack of a random sample and a lack of proper sample size and methodology.

I clearly understand the value of random sampling in a study. If you are trying to find out whether children do better in a straight marriage rather than a gay marriage, then you randomly sample straight marriages and gay marriages. This is what has been done numberous times by the APA with results showing that there is no significant difference.

When your random sample is to compare any relationship where someone who is gay with relationships where for the most part marriage is not allowed by our government with straight relationships where marriage is both allowed and encouraged, then you no longer have a random sample of similar situations. That is the fatal flaw in this study.

Perhaps at best, this study reinforces the importance of encouraging gays to marry and thereby having much stronger family relationships. To argue otherwise would be to be saying that marriage is not important in the stability and growth of children.

[Zeta-Flux]

I clearly understand the value of random sampling in a study. If you are trying to find out whether children do better in a straight marriage rather than a gay marriage, then you randomly sample straight marriages and gay marriages. This is what has been done numberous times by the APA with results showing that there is no significant difference.

I request evidence that the APA results did random samples. It is my understanding that is not the case; and in fact it would be *impossible* to do a random study on the groups you suggest because it would cost too much. (Because the groups are a very small minority of our population. Most gay people do not adopt, and of those that do, not all want to marry, or want to adopt because they had a kid in a previous marriage so it is a broken home, etc...)

When your random sample is to compare any relationship where someone who is gay with relationships where for the most part marriage is not allowed by our government with straight relationships where marriage is both allowed and encouraged, then you no longer have a random sample of similar situations. That is the fatal flaw in this study.

That is incorrect. You still have a random sample, it just does not consist of anyone in a SSM. With more states no allowing that, there will be more studies in the future which do measure that *new* group. In other words, it is a little strange to claim that the non-existence of the group 10 years ago is a *flaw* of the present study, unless you don't know what the study actually says.

Perhaps at best, this study reinforces the importance of encouraging gays to marry and thereby having much stronger family relationships. To argue otherwise would be to be saying that marriage is not important in the stability and growth of children.

The result says nothing of the kind, nor the opposite. It makes no causal claims.

[thesometimesaint]

http://www.apa.org/monitor/2010/10/same-sex.aspx

http://www.apa.org/monitor/2010/10/adopted-children.aspx

[Zeta-Flux]

thesometimesaint,

I clicked on the last link. It studies only 50 heterosexual couples and 56 homosexual couples. Definitely not a good sample size. Further, it was definitely not a random sample. (You don't get over 50% of the couples being homosexual in a random sample.)

I realize that part of my point may not be getting across because I'm being to brief in what I'm saying, so I'm going to try and write a longer post in the next couple hours.

Cheers,

Zeta-Flux

P.S. And notice the end of the blurb "With an estimated 100,000 children waiting to be adopted in the United States as of 2008, the research indicates that policies in states that forbid gay and lesbian couples from adopting need to be re-examined, Patterson said." You cannot get more political than that--straight from the study's author. Trying to make a political point after a small study from self-selected participants...good grief.

[Libs]

Politics should not even enter into the decision as to whether or not SSA can adopt. That should be a case by case decision, based on an applicant's suitability to parent..not on a person's sexual orientation. Taking a person's sexual orientation into consideration would be like taking a person's skin color into consideration.

A small sampling of 50 is more legitimate than a huge sampling of people who don't fit the profile of what it is you're trying to measure. A small, representative sample, of what you're trying to measure, is going to be more accurate, than a large random sampling that, even the researcher claims, doesn't really measure anything or mean anything, in terms of judging parenting.

[Zeta-Flux]

Okay, so I wanted to explain my ideas a little more fully, in the hopes that we can come to an agreement about what the Regnerus study does say, what future studies should do, and how it is easy to speak past one another.

First, I should make it clear that I think about anyone can make good parents. People who have been divorced. People who have been convicted of crimes. People who are natural birth-parents. People who adopt. The Regnerus study does not show that any particular person will be a bad parent. Moreover, it does not show that any specific class of parents will on average be bad parents. The only thing it purports to measure are the self-reported outcomes from a large random sample of the population, who are over 18 years old. It turns out that, for people in that group, those who experienced a parent who ever had a same-sex partner also self-reported a large number of negative outcomes. The study does not purport to claim a reason for this correlation. There could be a number of reasons: (1) the group was overly broad, and should have included only those who reported homosexual parents (i.e. no bisexualiy), (2) the fact that homosexual behavior was frowned upon for most of the lifetime of anyone 18 or older may have affected the issue, (3) the group also consisted of mostly broken homes--it is unclear whether that contributed to the problem or the homosexuality contributed to the broken homes, (4) homosexuality may, on average, in fact be detrimental, (5) something else, etc..., etc... To account for each of these theories, one would have to do a follow-up study.

Second, one of the common complaints about the study is that it will be misused. While this is true of all studies, there are a few measures that can be applied to see if the authors are contributing to this problem. (A) Does the author make it political? In this case the resounding answer is no. Unfortunately, the same is not true of other studies (just see the links above by thesometimesaint). (B) Is there a conflict of interest? In this case, there might be such, given the nature of the funding agency. But, to go any further, one would need to show that it is unusual to get funding from groups who have specific agendas. As far as I can tell, outside funding is not uncommon, there is no concrete reason to suppose Regnerus fudged the data in any way, and in fact the results of the actual study say almost nothing political (except when they say that it should not be used for political reasons; and neither should the other studies which were shown to be flawed). In fact, the same cannot be said of many of the studies in the other direction: they similarly get funding from pro-SSM groups, the data is not made public, the sample sizes are incredibly small and self-selected, and the authors of the study themselves make the studies political (by, as in the link provided by TSS, making political claims!!).

Third, another complaint against the Regnerus study is that it compares apples to fruit-cakes; stable homes to broken ones. This is a valid complaint if that were all the study did. Except that it also does compare broken homosexual homes to broken heterosexual homes.

Fourth, the author does admit a limitation in studying stable homosexual homes, simply due to numbers. Consider the following: About 2% of the population in the US claim to be homosexual all their lives. As we are assuming stable couples -- any child interviewed from such a union would be adopted from birth (here I include in vitro, etc...) Let's estimate high, and assume that 50% of gay couples are stable the entire lifetime of such children. Let's also estimate high, and say about 50% of all stable homosexual couples adopt a child birth. Our country has a population of about 300 million. Of that population, let's estimate high and say that in about 20% of the states (by population) that gay couples can adopt, form relationships, etc.. This gives us a sample size of about 300,000 in the entire US. For the sample size to be valid (for some measures), we need more than 300 people. That corresponds to about 1/1000 of the total sample size in America. To do a random study, you would have to interview 300,000 people. And this number is based on over-estimates, and the current ability for people to enter SSM.

One might have more luck in other countries, but once again you need an astronomical number of interviews to get down to the group you want. But further, even if you get there, it isn't clear that you are measuring what you want to measure anymore. We are assuming that same-sex couples form stable couplings at high rates, and adopt at high rates, etc... These might not be the case (even in other countries where there is no stigma) for a number of reasons -- (1) the culture (either nation-wide, or in the same-sex couple community) is against marriage or encourages behavior which prevents stable relationships, (2) the stigma is still present, etc... etc.... Or it might go the other way, those in stable SSM's might be predominantly wealthy and white, etc... ,etc...

To put it another way, if we define our group as stable couples, who never molested a child, who have never been divorced, never used drugs, don't drink or smoke, and have only been homosexual their entire lives, you remove by definitional fiat all of the bad affects that could naturally arise in the segment you really want to measure. Some of these conditionals may be appropriate, and some not, depending on exactly what you want to measure.

Anyway, I hope that clears things up.

And by the way, I think that this has almost nothing to do with whether or not SSM should be legal. On the other hand, I think this is an important question to consider with regards to adoption.

[Zeta-Flux]

libs,

A small sampling of 50 is more legitimate than a huge sampling of people who don't fit the profile of what it is you're trying to measure. A small, representative sample, of what you're trying to measure, is going to be more accurate, than a large random sampling that, even the researcher claims, doesn't really measure anything or mean anything, in terms of judging parenting.

First, as you admit, Regnerus doesn't purport to measure parenting.

Second, the fact that other people do want Regnerus' study to say that does not legitimize the small sampling size of another study, which the author thereof politicized. The two issues are completely separate. The fact that others misuse the Regnerus study does not make a small sampling size of another study "more accurate". That study is still as inaccurate and flawed as ever.

[Libs]

Zeta, I can agree with some of what you said above, but you do not, necessarily, have to have a large sampling, in order for a study to be reliable. The method is actually much more important that the number in the sample.

[selek1]

Zeta, I can agree with some of what you said above, but you do not, necessarily, have to have a large sampling, in order for a study to be reliable. The method is actually much more important that the number in the sample.

Indeed. True- a comparatively small sample (above a certain statistical threshold) can be workable, if the sample is also representative.

The sample sizes used in the earlier studies, however, were not only small, but they were either self-selecting or otherwise unrepresentative of the population as a whole.

Both flaws are almost invariably fatal to scientific accuracy.

[Libs]

Indeed. True- a comparatively small sample (above a certain statistical threshold) can be workable, if the sample is also representative.

The sample sizes used in the earlier studies, however, were not only small, but they were either self-selecting or otherwise unrepresentative of the population as a whole.

Both flaws are almost invariably fatal to scientific accuracy.

Not that I'm aware of, selek. They may have been carefully selected, since there were few married gay couples to choose from, but that's a positive, not a negative. This other study is flawed, in part, because it was too random and not selective enough. (By flawed, I mean, it doesn't show what some conservatives think it does)..

[Zeta-Flux]

libs,

Hey, even a little agreement is worth the time I spent. Cheers.

[Libs]

I appreciate the time you took to explain your position, Zeta. My posting here has been kind of hit and miss and I haven't put much time into it (apologies for that - been kinda busy).

I was very glad to hear that you believe anyone can be a good parent, regardless of orientation. We can certainly agree on that. As I've said, a couple of times, now, when it comes to suitability as parents, looking to adopt, I would really like to see the criteria based on an individual's character, and not on inborn traits like race or sexual orientation. In my view, that is the most fair approach.

Thanks again for your post.

[thesometimesaint]

Zeta:

It's getting late here so I'll respond in the morning. Good Night all.

[thesometimesaint]

Zeta-Flux:

There no intrinsic percentage that defines a "valid" representation. The average American is a white, middle class, lower middle aged women, living just outside of Kansas City; Kansas.

[Zeta-Flux]

Marks (2012) claims that samples with fewer than 393 respondents lack the necessary statistical power to detect small differences in outcomes. Here is a link to the Marks paper, which describes a large number of other problems with the APA studies. And here is a review by Cynthia Osborne discussing the Regnerus and Marks studies.

TSS, I just can't take your claim seriously. I'll admit quickly that I'm not an expert in these fields, but it seems like you just made those claims up whole-cloth, and they are not even slightly accurate. You might start by reading up on the differences between "average", "median", "mean", and "mode". Claiming that the "average American" is a white, middle class lower middle-aged female, is nonsensical. And if a study is limited to people only in that group, it definitely doesn't represent a study on the average situation. (Besides which, you didn't even address the bigger issue of the author thereof politicizing the results.)

[thesometimesaint]

zeta-flux:

average= total of all points divided by number of points. IE; 2+3+4=9 giving an average of 3.

median=midpoint of points. IE; 2,3,4 median is 3

mean=average

mode=most common point. IE; 2,3,3,4 mode is 3

Then you don't know a darn thing about statistical average.

I have no political agenda in the SS marriage debate. My ONLY point is that it is a best premature to make grandiose negative claims about the children of SS couples.

[Zeta-Flux]

My ONLY point is that it is a best premature to make grandiose negative claims about the children of SS couples.

There no intrinsic percentage that defines a "valid" representation.

TSS, who made grandiose negative claims? Who said anything about intrinsic percentages which define valid representations? All I claimed was that the study you linked was made overly political by the author (and it was), didn't have a large enough sample size to justify the claims (and I've now given a citation with evidence for this claim), and that the study was not done on a random, or even average, group of people.

You seem to understand averages, so take the population in American. Is the average person white? No, the average (if you define it as the sum of all data points divided by the total) is a mix. Same with gender. Age, you might be on to something.

When talking about groups of people, the average group is half male, half female. It is composed of a certain percentage of black, white, Hispanic, Asian, etc... It contains people of all sorts of ages (and not a homogeneous age). etc... This is why random sampling is so important; and to avoid statistical falsehoods this is why large enough groups are important. (Else, you get beans thrown at a map, as californiaboy put it.)

Then you don't know a darn thing about statistical average.

I am only a lowly professor of mathematics. You might be right.

Then again, my point that a study might be off if it only samples a self-selected group of people who are homogeneously white, educated, and financially well-off, may hold after all.

[thesometimesaint]

zeta-flux:

Did you not say this; "For the sample size to be valid (for some measures), we need more than 300 people. That corresponds to about 1/1000 of the total sample size in America. To do a random study, you would have to interview 300,000 people. And this number is based on over-estimates, and the current ability for people to enter SSM".

There is no arbitrary percentage needed for a valid measure. ALL survey's have a plus/minus factor. In theory it is possible to determine the average characteristics of a population by observing just one individual, however fraught with potential errors it may be.

The average person is not hermaphroditic. That is like saying the average couple has 2.4 children. No one has .4 children.

I know why we shoot for as large a representative study population as possible. The bigger the group the less likelihood of the "All Indians walk in single file. I saw one do it" problem".

With the exception of one hotly contested study there is no other study that concludes that the children of SS couples are measurably less well adapted than the children of different sex couples. It just isn't there as far as we can find. That doesn't necessarily mean that there IS no difference. Just that we have not been able to tease out what, if any difference there is so far. I personally suspect there is. But I have done no clinical studies on the subject to say definitively either way.

Technically speaking we're all an admixture of differing races, but as there is no scientific definition of race. The presenting characteristic of the average American is Caucasian.

I'm only a retired psychiatric social worker with a working lifetime in analyzing studies for methodological errors, and erroneous or premature conclusions.

Until we find a way to avoid the self selection trait we have to allow for it. But I know of no way to entirely avoid it. That will take bigger brains than mine.

[selek1]

Not that I'm aware of, selek.

That lack of awareness is demonstrably attributable to willful ignorance.

You have been provided with a number of references showing exactly that- and you've either ignored those references or dismissed them out of hand.

They may have been carefully selected, since there were few married gay couples to choose from, but that's a positive, not a negative

By definition, a "selected" sample of any kind is not a "random" sample. Worse, the particpants were self-selected, based (IIRC) on their frequenting certain "gay-friendly" businesses.

This is not in dispute- even University of California-Davis psychology professor Gregory Herek has admitted that Marks criticisms of the APA paper are accurate.

Being both "small" and "self-selecting" it cannot be reasonably argued to be a representative sample.

If you poll 100 homosexual men in downtown Berkeley, your results will be skewed liberal-left and pro-homosexual.

If you poll 100 heterosexual men in Austin Texas, your results will be skewed conservative-right and anti-homosexual.

Neither will give you an accurate representation of the country at large.

As Marks has shown- and as others have repeatedly brought to your attention- the APA studies used small, homogenous, self-selecting sample that skewed the results. Despite those flaws, you and your fellow travellers stood up and cheered results that confirmed your biases.

Regnerus, by contrast, used a large, random sample to correct for that flaw- and he has been pilloried for it.

My point- then and now- is that all of your pious and pseudo-intellectual posturing about "scientific accuracy" and "rigorous standards" goes out the window when the study agrees with your pre-determined conclusions.

And that's not science. That's dogma.

This other study is flawed, in part, because it was too random and not selective enough.

Call For References- how can a finite sample be too random?

The Regnerus study boasts a random sample that is statistically large enough to provide an accurate reflection of the population at large. It controlled for a large number of factors, including age, race, gender, and things like being bullied as a youth, or the "gay-friendliness" of the of the state in which the respondents lived.

(By flawed, I mean, it doesn't show what some conservatives think it does)..

In other words, as has been repeatedly stated, it is not the study's methodolgy that bothers you (otherwise, you'd be just as swift to condemn the APA study), but the fact that it is being used to savage your sacred cow.

[selek1]

The bottom line is this- and all the desperate flailing about cannot conceal this fundamental truth.

Libs has personally demonstrated this truth in post after post.

If the Regnerus study had shown that children perform just as well in homosexual homes as they do in heterosexual homes, then none of the critics here would care about the alleged flaws of the study.

Their antipathy to Regnerus and hostility to his study are driven solely by the impact the fear his work will have on advancing their ideology.

Marks and Regnerus have committed the penultimate sin of impeding the path to Utopia- and must therefore be vilified, dismissed, and destroyed- and intellectual honesty be damned.